The Case for Melding Engagement and Sentiment
Walk into any modern newsroom today and you will find a desk for the audience, growth, or engagement team. Walk to that desk, and you will most likely find a team of these editors crowing about one of these two things: engagement and sentiment. Digging deeper, these two words aren’t easily defined by the people who use them. Giselle Abramovich’s Digiday post outlines the quandary:
“Engagement and sentiment may not be the best bet for judging the effectiveness of social media efforts. What is “engagement” anyway? The definition would be different for all brands. For a link that, let’s say, Adobe posts on its Facebook page for a free trial of Photoshop, engagement would mean clicking on the link and downloading the trial. But for a video that a brand like Club Monaco posts on Facebook, engagement may mean video likes, views, shares and comments.”
An Ad Age article outlines the same frustration, saying “if you ask 100 different marketers to define engagement, you’ll get almost as many answers.” From clickthrough rates and likes, shares, and comments, to video views and impressions, a universal definition for engagement is impossible.
The same goes for “sentiment.” Few publications have ventured to define it, opting instead to outline what analyzing it means. Nieman Lab defines sentiment as “what a percentage of the population “feels” about something.” Simply Measured defines sentiment analysis as “detecting and understanding how the audience is reacting to a brand, either positively or negatively.” Adobe goes even further, listing the following as key performance indicators for sentiment:
- Amplification rate — tracks the rate at which your audience actively shares your content through their social networks
- Applause rate — reveals the rate at which your social community provides positive feedback through shares, +1s, and likes
- Audience growth rate — measures the rate of social network growth over a period of time
- Average engagement rate — tracks the average engagement against your follower base
- Visitor frequency rate — compares new and returning followers
Sam Petulla of Nieman goes on to say that much of the analysis is done via “machine learning” or “affective learning,” which are fancy ways of saying that a computer reads millions of reactions to content and grades it based on its positivity or negativity. The site has published numerous articles about platforms and startups that attempt to use these new technologies to predict sentiment. The same goes for engagement, with many companies like Adobe and IBM analyzing publication and social media metrics for publishers and brands. These companies, though, are being challenged by a crop of newer companies attempting to bridge sentiment and engagement.
Pollware by Sodahead helps publishers generate revenue with custom polling. Whatsgoodly bills itself as a database of millennial opinions, decided by multiple choice questions curated via college campuses; the firm then uses the insights to create “stories.” The Peal curates and scores political content based on an upvote-downvote system. Reddit, the front page of the Internet, takes that a step further by allowing users to create their own communities in addition to an upvote-downvote system. Finally, Playbuzz tries to mash together engagement and sentiment by having a host of custom quizzes and polls for both users and publishers.
While the breadth of technology companies and pundits trying to put a finger on engagement and sentiment is refreshing, it seems like they are missing a connecting thread. They are all tugging at the same strings, but none at all of them. While Reddit assigns a score based on how many users upvote a story or piece of content, there is no relationship established between content. While Whatsgoodly and The Peal attempt to segment opinions and content by subject, they leave a lot to be desired in terms of breadth. Playbuzz, the most wide-ranging of the services, is almost too broad in the amount of quizzes, polls, and options it offers.
That leaves a large hole in the market for a platform that not only measures engagement and sentiment, but also establishes a relationship between similar and disparate content and allows for segmentation that breeds community. It leaves a hole for my project: Flickie.
Flickie turns the simple gesture of versus-style swiping into an engagement and sentiment measurement for publishers, brands, and content consumers. It leverages the act of ‘flicking’ left or right on a story, a piece of content, or a subject to create one number that everyone can agree on as a measure of what people “feel” about it. Publishers and newspapers can use it to measure whether readers like a story, whether they like the actors or subject in the story, and then to promote their other related stories. People can use it to rank their favorite content and stories and discover new content and stories. It will meld engagement and sentiment in a way that lends itself to discussion and discourse. In short, it will make the subjective objective across the publishing and social spheres.
